Arxiv:1910.09700 Autotrain compatible Conversational Endpoints compatible Moe Qwen2 moe Region:us Safetensors Sharded Tensorflow
Model Card on HF 🤗: https://huggingface.co/khoantap/cheap-moe-merge
Cheap MoE Merge Benchmarks
MMLU Pro: 25.98
GPQA: 6.04
MUSR: 13.32
BBH: 29.8
IFEval: 45.57 vs 88 (so35)-48.2%
MATH Lvl 5: 9.21
LLME Score: 0.26
nn.n% — How the model compares to the reference models: Anthropic Sonnet 3.5 ("so35"), GPT-4o ("gpt4o") or GPT-4 ("gpt4").
What is the LLM Explorer Rank (Score)
Cheap MoE Merge Parameters and Internals
| LLM Name | Cheap MoE Merge |
| Repository 🤗 | https://huggingface.co/khoantap/cheap-moe-merge |
| Model Size | 19.3b |
| Required VRAM | 38.5 GB |
| Updated | 2025-09-16 |
| Maintainer | khoantap |
| Model Type | qwen2_moe |
| Model Files | |
| Model Architecture | Qwen2MoeForCausalLM |
| Context Length | 131072 |
| Model Max Length | 131072 |
| Transformers Version | 4.46.3 |
| Tokenizer Class | PreTrainedTokenizerFast |
| Padding Token | <|begin_of_text|> |
| Vocabulary Size | 128256 |
| Torch Data Type | float16 |
Best Alternatives to Cheap MoE Merge
Best Alternatives |
Context / RAM |
Downloads |
Likes |
|---|---|---|---|
| MoE Out Merge | 128K / 38.5 GB | 6 | 0 |
Rank the Cheap MoE Merge Capabilities
🆘 Have you tried this model? Rate its performance. This feedback would greatly assist ML community in identifying the most suitable model for their needs. Your contribution really does make a difference! 🌟
What open-source LLMs or SLMs are you in search of? 51545 in total.
Original data from HuggingFace, OpenCompass and various public git repos.
Release v20241124